I'm using the quantmod package to import financial series data from Yahoo.
library(quantmod)
getSymbols("^GSPC")
[1] "GSPC"
I'd like to change the name of object "GSPC" to "SPX". I've tried the rename function in the reshape package, but it only changes the variable names. The "GSPC" object has vectors GSPC.Open, GSPC.High, etc. I'd like my renaming of "GSPC" to "SPX" to also change GSPC.Open to SPX.Open and so on.
Renaming an object and the colnames within it is a two step process:
SPY <- GSPC # assign the object to the new name (creates a copy)
colnames(SPY) <- gsub("GSPC", "SPY", colnames(SPY)) # rename the column names
Otherwise, the getSymbols function allows you to not auto assign, in which case you could skip the first step (you will still need to rename the columns).
SPY <- getSymbols("^GSPC", auto.assign=FALSE)
Comment from @backlin
R employs so-called lazy evaluation. An effect of that is that when you "copy" SPY <- GSPC
you do not actually allocate new space in the memory for SPY
. R knows the objects are identical and only makes a new copy in the memory if one of them is modified (i.e. when they are no longer the identical, e.g. when you change the column names on the following line). So by doing
SPY <- GSPC
rm(GSPC)
colnames(SPY) <- gsub("GSPC", "SPY", colnames(SPY))
you never really copy GSPC
but merely give it a new name (SPY
) and then tell R to forget the first name (GSPC
). When you then change the column names you do not need to create a new copy of SPY
since GSPC
no longer exists, meaning you have truly renamed the object without creating intermediate copies.